Computers, materials & continua/Computers, materials & continua (Print),
Journal Year:
2024,
Volume and Issue:
80(1), P. 1601 - 1631
Published: Jan. 1, 2024
Virtual
machine
(VM)
consolidation
aims
to
run
VMs
on
the
least
number
of
physical
machines
(PMs).
The
optimal
significantly
reduces
energy
consumption
(EC),
quality
service
(QoS)
in
applications,
and
resource
utilization.
This
paper
proposes
a
prediction-based
multi-objective
VM
approach
search
for
best
mapping
between
PMs
with
good
timeliness
practical
value.
We
use
hybrid
model
based
Auto-Regressive
Integrated
Moving
Average
(ARIMA)
Support
Vector
Regression
(SVR)
(HPAS)
as
prediction
consolidate
results
by
HPAS,
aiming
at
minimizing
total
EC,
performance
degradation
(PD),
migration
cost
(MC)
wastage
(RW)
simultaneously.
Experimental
using
Microsoft
Azure
trace
show
proposed
has
better
accuracy
overcomes
without
(i.e.,
Non-dominated
sorting
genetic
algorithm
2,
Nsga2)
renowned
Overload
Host
Detection
(OHD)
approaches
prediction,
such
Linear
(LR),
Median
Absolute
Deviation
(MAD)
Inter-Quartile
Range
(IQR).
IGI Global eBooks,
Journal Year:
2025,
Volume and Issue:
unknown, P. 177 - 200
Published: Feb. 5, 2025
In
today's
digital
world,
managing
large
volumes
of
data,
known
as
“Big
Data,”
presents
a
significant
challenge
due
to
its
volume
and
complexity.
Regular
software
often
struggles
handle
this,
necessitating
the
use
Load
Balancing—a
crucial
aspect
cloud
computing.
balancing
distributes
workloads
across
resources,
preventing
slowdowns,
reducing
processing
time,
optimizing
system
performance.
This
paper
explores
load
strategies
in
big
data
processing,
including
Round
Robin,
Least
Connection,
Resource-Based,
Task-Based,
Dynamic
methods,
discussing
their
pros
cons.
Effective
ensures
optimal
resource
usage,
higher
scalability,
increased
availability,
dependability,
fault
tolerance,
improved
The
provides
literature
review,
proposes
model
for
balancing,
tests
it
simulated
environment,
highlighting
key
findings
suggesting
future
research
directions.
Applied Sciences,
Journal Year:
2021,
Volume and Issue:
11(13), P. 5849 - 5849
Published: June 23, 2021
Cloud
computing
is
a
rapidly
growing
technology
that
has
been
implemented
in
various
fields
recent
years,
such
as
business,
research,
industry,
and
computing.
provides
different
services
over
the
internet,
thus
eliminating
need
for
personalized
hardware
other
resources.
environments
face
some
challenges
terms
of
resource
utilization,
energy
efficiency,
heterogeneous
resources,
etc.
Tasks
scheduling
virtual
machines
(VMs)
are
used
consolidation
techniques
order
to
tackle
these
issues.
extensively
studied
literature.
The
problem
with
parameters
objectives.
In
this
article,
we
address
consumption
efficient
utilization
virtualized
cloud
data
centers.
proposed
algorithm
based
on
task
classification
thresholds
better
utilization.
first
phase,
workflow
tasks
pre-processed
avoid
bottlenecks
by
placing
more
dependencies
long
execution
times
separate
queues.
next
step,
classified
intensities
required
Finally,
Particle
Swarm
Optimization
(PSO)
select
best
schedules.
Experiments
were
performed
validate
technique.
Comparative
results
obtained
benchmark
datasets
presented.
show
effectiveness
algorithms
which
it
was
compared
consumption,
makespan,
load
balancing.
